UAV Thermal Images for Water Presence Detection in a Mediterranean Headwater Catchment

As Mediterranean streams are highly dynamic, reconstructing space–time water presence in such systems is particularly important for understanding the expansion and contraction phases of the flowing network and the related hydro–ecological processes. Unmanned aerial vehicles (UAVs) can support such monitoring when wide or inaccessible areas are investigated. In this study, an innovative method for water presence detection in the river network based on UAV thermal infrared remote sensing (TIR) images supported by RGB images is evaluated using data gathered in a representative catchment located in Southern Italy. Fourteen flights were performed at different times of the day in three periods, namely, October 2019, February 2020, and July 2020, at two different heights leading to ground sample distances (GSD) of 2 cm and 5 cm. A simple methodology that relies on the analysis of raw data without any calibration is proposed. The method is based on the identification of the thermal signature of water and other land surface elements targeted by the TIR sensor using specific control matrices in the image. Regardless of the GSD, the proposed methodology allows active stream identification under weather conditions that favor sufficient drying and heating of the surrounding bare soil and vegetation. In the surveys performed, ideal conditions for unambiguous water detection in the river network were found with air–water thermal differences higher than 5 °C and accumulated reference evapotranspiration before the survey time of at least 2.4 mm. Such conditions were not found during cold season surveys, which provided many false water pixel detections, even though allowing the extraction of useful information. The results achieved led to the definition of tailored strategies for flight scheduling with different levels of complexity, the simplest of them based on choosing early afternoon as the survey time. Overall, the method proved to be effective, at the same time allowing simplified monitoring with only TIR and RGB images, avoiding any photogrammetric processes, and minimizing postprocessing efforts.

A scSE-LinkNet Deep Learning Model for Daytime Sea Fog Detection

Sea fog is a precarious weather disaster affecting transportation on the sea. The accuracy of the threshold method for sea fog detection is limited by time and region. In comparison, the deep learning method learns features of objects through different network layers and can therefore accurately extract fog data and is less affected by temporal and spatial factors. This study proposes a scSE-LinkNet model for daytime sea fog detection that leverages residual blocks to encoder feature maps and attention module to learn the features of sea fog data by considering spectral and spatial information of nodes. With the help of satellite radar data from Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP), a ground sample database was extracted from Moderate Resolution Imaging Spectroradiometer (MODIS) L1B data. The scSE-LinkNet was trained on the training set, and quantitative evaluation was performed on the test set. Results showed the probability of detection (POD), false alarm rate (FAR), critical success index (CSI), and Heidke skill scores (HSS) were 0.924, 0.143, 0.800, and 0.864, respectively. Compared with other neural networks (FCN, U-Net, and LinkNet), the CSI of scSE-LinkNet was improved, with a maximum increase of nearly 8%. Moreover, the sea fog detection results were consistent with the measured data and CALIOP products.

A Double Swath Configuration for Improving Throughput and Accuracy of Trait Estimate from UAV Images

Multispectral observations from unmanned aerial vehicles (UAVs) are currently used for precision agriculture and crop phenotyping applications to monitor a series of traits allowing the characterization of the vegetation status. However, the limited autonomy of UAVs makes the completion of flights difficult when sampling large areas. Increasing the throughput of data acquisition while not degrading the ground sample distance (GSD) is, therefore, a critical issue to be solved. We propose here a new image acquisition configuration based on the combination of two focal length (f) optics: an optics with f=4.2 mm is added to the standard f=8 mm (SS: single swath) of the multispectral camera (DS: double swath, double of the standard one). Two flights were completed consecutively in 2018 over a maize field using the AIRPHEN multispectral camera at 52 m altitude. The DS flight plan was designed to get 80% overlap with the 4.2 mm optics, while the SS one was designed to get 80% overlap with the 8 mm optics. As a result, the time required to cover the same area is halved for the DS as compared to the SS. The georeferencing accuracy was improved for the DS configuration, particularly for the Z dimension due to the larger view angles available with the small focal length optics. Application to plant height estimates demonstrates that the DS configuration provides similar results as the SS one. However, for both the DS and SS configurations, degrading the quality level used to generate the 3D point cloud significantly decreases the plant height estimates.

Identification of Marine Debris and Its Distribution Using Unmanned Aerial Vehicle (UAV) on the Cirebon Coastal Area, Indonesia

Marine debris can be a significant problem when it enters the ocean. One of the areas, which has a marine debris problem is Cirebon Coast. Such a problem occurs due to the high human population and activities in this particular coastal area. An effective cleaning method is required to prevent this problem. However, to determine the cleaning method, comprehensive information about the debris condition is required. Therefore, this study aims to identify the dominant types of marine debris and analyze the effect of tides on the characteristics and distribution of debris on the Cirebon Coast. This study carried out marine debris identification by orthophoto obtained from a DJI Phantom 4 Pro Unmanned Aerial Vehicle (UAV)/drone. The dominant types of marine debris on the Cirebon Coast are plastic and styrofoam. Based on the aerial imagery validation results, plastic and styrofoam larger than Ground Sample Distance (GSD) are easily detected. Visually, debris in Karang Anom more than in Rawa Urip. The change in tidal height can affect debris not visible on the orthophoto at the maximum water level. In addition, the tides can also move marine debris varying from 10 to 50 cm from its previous position. The debris area at Rawa Urip Beach when slack before flood tide (55.53 m2) is larger than the area when slack before ebb tide (52.71 m2). The debris area at Karang Anom Beach at low tide (129.89 m2) is larger than when slack before ebb tide (75.79 m2). This study revelaed that the factors affecting debris area on the Cirebon Coast comprise tidal height, seawater visibility, and the beach structure (slope and coast components).

Improving flood damage assessments in data-scarce areas by retrieval of building characteristics through UAV image segmentation and machine learning – a case study of the 2019 floods in southern Malawi

Reliable information on building stock and its vulnerability is important for understanding societal exposure to floods. Unfortunately, developing countries have less access to and availability of this information. Therefore, calculations for flood damage assessments have to use the scarce information available, often aggregated on a national or district level. This study aims to improve current assessments of flood damage by extracting individual building characteristics and estimate damage based on the buildings' vulnerability. We carry out an object-based image analysis (OBIA) of high-resolution (11 cm ground sample distance) unmanned aerial vehicle (UAV) imagery to outline building footprints. We then use a support vector machine learning algorithm to classify the delineated buildings. We combine this information with local depth–damage curves to estimate the economic damage for three villages affected by the 2019 January river floods in the southern Shire Basin in Malawi and compare this to a conventional, pixel-based approach using aggregated land use to denote exposure. The flood extent is obtained from satellite imagery (Sentinel-1) and corresponding water depths determined by combining this with elevation data. The results show that OBIA results in building footprints much closer to OpenStreetMap data, in which the pixel-based approach tends to overestimate. Correspondingly, the estimated total damage from the OBIA is lower (EUR 10 140) compared to the pixel-based approach (EUR 15 782). A sensitivity analysis illustrates that uncertainty in the derived damage curves is larger than in the hazard or exposure data. This research highlights the potential for detailed and local damage assessments using UAV imagery to determine exposure and vulnerability in flood damage and risk assessments in data-poor regions.

Novel Permanent Magnetic Surface Work Hardening Process for 60/40 Brass

Surface work hardening is a process of deforming a material surface using a thin layer. It hardens and strengthens the surface while keeping the core relatively soft and ductile to absorb stresses. This study introduces a permanent magnate surface work hardening under two opposite permanent poles of a magnet to investigate its influence on a brass surface. The gap between the brass and the north magnet pole—fixed in the spindle of a vertical machine—was filled with martensitic stainless steel balls. The rotational speed and feed rates were 500–1250 rpm and 6–14 mm min−1, respectively. The novel method improved the surface hardness for all parameters by up to 112%, in favor of high speed, and also increased yield by approximately 10% compared to ground samples. Surface roughness showed higher values for all speed–feed rate combinations compared to the ground sample. Nevertheless, it showed better roughness than other treated conditions with high and low feed rates. The ultimate tensile strength and ductility remained unchanged for all conditions other than the untreated brass. A factorial design and nonlinear regression analysis were performed to predict the microhardness equation and effectiveness of the independent variable—speed and feed rate—for the proposed process.

Steps to build a DIY low-cost fixed-wing drone for biodiversity conservation

Despite the proved usefulness of drones in biodiversity studies, acquisition costs and difficulties in operating, maintaining and repairing these systems constrain their integration in conservation projects, particularly for low-income countries. Here we present the steps necessary to build a low-cost fixed-wing drone for environmental applications in large areas, along with instructions to increase the reliability of the system and testing its performance. Inspired by DIY (Do It Yourself) and open source models, this work prioritizes simplicity and accounts for cost-benefit for the researcher. The DIY fixed-wing drone developed has electric propulsion, can perform pre-programmed flight, can carry up to 500 g payload capacity with 65 minutes flight duration and flies at a maximum distance of 20 km. It is equipped with a RGB (Red, Green and Blue) sensor capable of obtaining 2.8 cm per pixel Ground Sample Distance (GSD) resolution at a constant altitude of 100 m above ground level (AGL). The total cost was $995 which is substantially less than the average value of similar commercial drones used in biodiversity studies. We performed 12 flight tests in auto mode using the developed model in protected areas in Brazil, obtaining RGB images that allowed us to identify deforestation spots smaller than 5 m2 and medium-sized animals. Building DIY drones requires some technical knowledge and demands more time than buying a commercial ready-to-fly system, but as proved here, it can be less expensive, which is often crucial in conservation projects.

Parameterization of the Individual Tree Detection Method Using Large Dataset from Ground Sample Plots and Airborne Laser Scanning for Stands Inventory in Coniferous Forest

Highly accurate and extensive datasets are needed for the practical implementation of precision forestry as a method of forest ecosystem management. Proper processing of huge datasets involves the necessity of the appropriate selection of methods for their analysis and optimization. In this paper, we propose a concept for and implementation of a data preprocessing algorithm, and a method for the empirical verification of selected individual tree detection (ITD) algorithms, based on Airborne Laser Scanning (ALS) data. In our study, we used ALS data and very extensive dendrometric field measurements (including over 21,000 trees on 522 circular sample plots) in the economic and protective coniferous stands of north-eastern Poland. Our algorithm deals well with the overestimation problems of tree top detection. Furthermore, we analyzed segmentation parameters for the two currently dominant ITD methods: Watershed (WS) and Local Maximum Filter with Growing Region (LMF+GR). We optimized them with respect to minimizing the Root Mean Square Error (RMSE) and Mean Absolute Percentage Error (MAPE). Additionally, our results show the crucial importance of the quality of empirical data for the correct evaluation of the accuracy of ITD algorithms.

Seed Rain Potential in Late-season Weed Escapes can be Estimated Using Remote Sensing

The presence of a soil seedbank facilitates the persistence of annual weed species in arable fields. Soil weed seedbank is replenished by many sources, but the largest one is the seeds produced by uncontrolled weed escapes present during late-season. The estimation of weed seed production potential from late-season escapes may allow farmers to make appropriate management decisions to minimize seedbank replenishment. The objective of this research was to evaluate the feasibility of using UAV-based RGB and multispectral imagery for estimating seed rain potential in late-season weed escapes in crop fields. Three case studies were used to capture images of weed escapes prior to crop harvest: waterhemp [Amaranthus tuberculatus (Moq.) Sauer] in soybean [Glycine max (L.) Merr.], Palmer amaranth [Amaranthus palmeri (S.) Watson] in cotton (Gossypium hirsutum L.), and johnsongrass [Sorghum halepense (L.) Pers.] in soybean. Randomly selected quadrats with different density gradients of weed escapes were sampled at the time of crop maturity. High-resolution RGB and multispectral images of the experimental area were collected using drones immediately prior to ground sample collection. Normalized difference vegetation index (NDVI), excess green index (ExG), and canopy volume estimates derived from canopy height models were used to obtain weed biological measurements (biomass and seed production). Among the indices investigated, NDVI and ExG had very strong correlations (0.71 – 0.97) with weed biomass. No specific remote sensing variable was ideal across the three cases examined here, suggesting that a generalized remote sensing approach may not offer robust estimations, and case-specific applications are imperative. Nonetheless, drone imagery is a powerful tool for estimating seed production from uncontrolled weed escapes and assist with management decision making.

Comprehensive Evaluation of the GF-4 Satellite Image Quality from 2015 to 2020

GaoFen-4(GF-4) is the highest spatial resolution Earth observation satellite operating in geosynchronous orbit. Its fixed Earth observation location, rapid responsiveness, and wide observation range make it popular in disaster and emergency monitoring. To evaluate the GF-4 image quality in detail on a long-term basis, this study analyzes the image quality after the commissioning phase by focusing on ground sample distance (GSD) and geometric and radiometric quality. The theoretical calculation, geometric and radiometric measurements, and on-site experiments results show that (1) the GSD of the GF-4 image is ~50 m at the nadir point and increases gradually with the distance away from the nadir point, (2) most external geometric errors are within the design requirements of 4 km despite some exceeding the limit, and the internal geometric errors are tested within 1 pixel, and (3) image sharpness is generally stable but varies with the atmosphere condition and imaging time, and the radiometric response gradually degrades at the rate of less than 5.5% per year.